By Tom Foremski, Silicon Valley Watcher

The best entrepeneurs know that it is important to think big, because big ideas excite and motivate people.

Applying this concept to the internet, I wondered what type of internet business could be started that had the potential to be the largest, and what kind of business would it be?

Would it be a larger Yahoo? Or maybe a mega online retailer combining an Amazon and Ebay?

In fact it is none of the above or any other combination of internet companies that you, or Wall Street can imagine.

The largest potential internet business would have to be the internet itself.

But how could a startup create a competitor to the internet? The capital costs of a new internet would be staggering, let alone the marketing, and all the rest.

Well, let's not get distracted by big investment numbers just yet. Let's consider what would be the most important characteristic of a successful new internet: It would have to be compellingly better and cheaper than the old internet, by several magnitudes preferably. And if we were designing a new internet today, it would look nothing like the patchwork of systems in use today.

The design of a new internet would mandate using the highest cost performance IT architecture available, which means industry standard hardware, open-source software components, and a fair amount of Big Iron--large computer systems to handle large computational problems more efficiently than clustered smaller systems.

And the design of a new internet would require special attention to electric power consumption because energy costs are now a large factor in cost/performance calculations. And those costs will increase as we move further from peak oil production.

Individuals, companies, countries, are all seeking to lessen exposure to energy prices and supplies. Because the consequences of being over-exposed raise costs in a world where competitive leadership is all about being the lowest cost/highest quality producer/provider. Not to mention the disruption to trade and society from wars over energy resources.

We know the internet is very important to our future, and to those in other countries too. It is a shared global platform that will gain, not lessen in importance. And therefore it should be powerful, incredibly reliable, and ubiquitous, and very low in energy consumption.

And such an internet could be built with today's components and technologies and be a boon to all users. It would not require the invention of any exotic technologies.

A new internet would vault digital divides and network the entire world population because the edges of the network could be accessed by inexpensive digital devices: digital phones, simple PCs and notebooks, using processing cycles within the internet.

Building a new internet could become a symbol for a global civilization based on collaboration and understanding. And that is what Google is doing, imho.